Session Timeline
You're in a War Room prepping for an impending launch. A stakeholder report just came in — there are bugs that need to be fixed ASAP.
Kickoff
Set the stage, walk through goals, mindset, and AI prompting.
Bug Blitz
Pick your role, form your War Room squad, and review the Bug Report from stakeholders.
Debrief
Review pillars and share your War Room experience. If time allows, we will review a bug together.
Setup
What to Bring
Required
- A laptop with a modern web browser
Optional (but helpful)
- A free-tier AI account (ChatGPT, Gemini, Claude, etc.)
- An AI coding assistant (Cursor, Windsurf, GitHub Copilot)
- Familiarity with browser DevTools
We Provide
- The Bug Blitz app with pre-baked bugs
- Workshop Guide
- Bug Report from stakeholders
- Cheat sheet with hints
Accessing the Codebase & Website
Launch the app in StackBlitz for a quick-start workspace with no installation required. If you prefer running locally, you can clone the repository from GitHub.
Launch the Workshop App
Launch in StackBlitz
Enter your table number (1–20) or scan the QR code at your table. This connects you to your team's isolated database schema.
⚠️ Important:
The ?team=<id> parameter routes database requests
to your squad's isolated schema in Supabase. Do not remove it.
Clone & Run Locally
Prerequisite: Node.js 22 or newer.
Check your version with node --version.
💻 Need to install or upgrade Node? Download it from nodejs.org, or use a version manager: nvm on macOS/Linux or nvm-windows on Windows.
Clone the repository from https://github.com/marisamachlis/bugblitz-ghc2026, then choose one of the paths below.
Recommended and works on any OS.
npm install && npm start
If you already have make and nvm,
the repo can handle setup and startup for you.
make setup && make dev
If make errors or is not set up, we recommend following Path A.
-
Via URL:
Navigate to
http://localhost:4200/?team=<table number> -
Via Code:
Set
team: '<table number>'insidesrc/environments/environment.ts
Explore the App
Browse the plant store — shop for plants, add items to cart, try checking out. Notice anything off? That's a bug. Use browser DevTools to inspect the page and view the console, network tools, and DOM.
Open DevTools with
F12,
Option + Command + I on Mac, or
Ctrl + Shift + I on Windows.
Read the Bug Report
Your team will receive a Bug Report with issues reported by stakeholders — some vague, some specific.
Prioritize & Fix
As a team, decide which bugs to tackle first based on severity, time constraints, and your role's perspective.
Workshop Goals
By the end of the hour, you'll have practiced four things that actually matter on release day.
Confidence Building
Walk into your next launch day knowing you have a repeatable playbook for the moment things break.
Teamwork & Communication
Advocate for priorities through your role's lens, share context, and land at a decision the whole War Room can back.
Triage Mastery
Sort a pile of incoming bug reports into P1/P2/P3 and level-of-effort — quickly, and with reasoning you can defend.
Debugging & DevTools
Practice using the browser Console, Elements, and Network tabs — plus AI as a debugging partner — to isolate root causes fast.
Bug-Busting Mindset
Bugs on release day come with the territory. A calm War Room isn't one where nothing broke — it's one that knows how to react. Three pillars carry that.
Psychological Safety
Reward Curiosity
Respond together, not with blame or panic. Surface what's broken and share context.
Strategic Triage & Trade-Offs
Functional over Flawless
Weigh impact against effort. Resolve true release blockers now; defer the rest.
Leveraging Your Voice
Own Your Reasoning
Back your decisions with clear rationale. Make sure your reasoning is visible, not just assumed.
Objectives
By the end of the workshop, every participant should have practiced these four moves.
- Triage First. Assign priority and effort to the reported bugs. Sort them into P1/P2/P3 with a quick effort estimate and be ready to defend your ranking in one sentence per ticket.
- Share the Why. Base decisions on user impact and dev effort — not gut feel. Your teammates should be able to follow your reasoning without asking.
- Choose Your Approach. Decide between a quick workaround and a comprehensive fix. Both are right answers in the right column of the Priority × Effort matrix.
- Problem Solve. Leverage DevTools (Console, Network, Elements) and AI agents — briefed, audited, owned — to actually squash the bugs.
Prompting
AI agents can accelerate your debugging — but only if you prompt them well. Know where they shine and where they need a push, then use the three-step framework below on every prompt.
AI's strong suits
- Decoding a cryptic error — paste the surrounding code and one sentence about what you were doing; you'll usually get a real lead.
- Spotting the mechanical stuff: typos, missing semicolons, swapped argument order. Anything a second read would catch.
- Drafting unit tests once you've told it the behavior you want. Coverage without the boilerplate.
Where AI needs a nudge
- It can't see your app. Describe what's on screen and what you expected instead — vague prompts get vague fixes.
- Don't accept the first suggestion. Make it defend its logic, then read the diff before you click accept.
- Feed it the framework, file paths, and data shape — otherwise you'll get textbook Angular answers to your very specific bug.
You can't throw a vague request at an agent and expect production-ready code. To get the most out of AI during a high-pressure release, use a three-step framework: Brief, Audit, and Own.
Brief the Agent
- Define your role and the specific issue.
- Provide full context — logs, code, environment.
- State the current behavior vs. the expected behavior.
Audit the Output
- Don't assume AI is correct — watch for hallucinations.
- Force the agent to defend its logic.
- Stress-test for missing edge cases.
Own the Solution
- Never just click "Accept."
- Ask for step-by-step breakdowns.
- Cross-reference with reputable documentation.